The magnetic Fe3O4 nanoparticles were synthesized according to the method described in the literature . Transmission electron microscope (TEM) images (Fig. S1a) showed that the Fe3O4 nanoparticles were uniform in size with a diameter of approximately 4-8 nm. Moreover, a clear lattice structure is shown, which indicates that it has single-crystal properties that were further demonstrated through X-ray diffraction characterization (XRD in Fig. S1b). All diffraction peaks for Fe3O4 can be indexed to the , , , , , and  planes, which is consistent with the standard values of Fe3O4 according to the references of JCPDS card no. JCPDS 19-0629 [33, 37–38]. The VSM measurements (Fig. S1c) showed that the saturation magnetization (Sm) of ferric oxide was 44.8 emu/g, and the remanence and coercivity were zero, which may be related to its small particle size.
Furthermore, a magnetic nanosensor of Fe3O4@Os1-PS was prepared through emulsifier polymerization with an oxygen probe Os1, polymeric monomer (styrene), and magnetic Fe3O4 nanoparticles, as shown in Fig. 1. The magnetic nanosensor of Fe3O4@Os1-PMMA was adopted using the same synthetization method as in the case of Fe3O4@Os1-PS, by changing the styrene for methyl methacrylate, as shown in Fig. S2.
The nanoparticles have gained sizes of 50−100 nm after 5 hours under the reaction with protection of nitrogen. In order to know the sizes and morphologies of two types of nanoparticles, transmission electron microscope (TEM) and scanning electron microscope (SEM) were detected as they are the important means to demonstrate. TEM and SEM images for Fe3O4@Os1-PS and Fe3O4@Os1-PMMA are shown in Fig. 2. As shown in the TEM images of Fe3O4@Os1-PS (Fig. 2a) and Fe3O4@Os1-PMMA (Fig. 2b), it is evident that the Fe3O4 nanoparticles were wrapped in nanoparticles polymerized with styrene or methyl methacrylate as monomers, respectively. From Fig. 2c and Fig. 2d, both of these nanoparticles had diameter sizes of approximately 80 nm which can also be proved by the characterization of Dynamic lighting scattering (DLS). DLS (Fig. S3a and Fig. S3b) showed excellent dispersibility with low PDI, with 0.198 for Fe3O4@Os1-PS and 0.202 for Fe3O4@Os1-PMMA.
To study the oxygen-sensing properties of Fe3O4@Os1-PS and Fe3O4@Os1-PMMA in aqueous solutions, the emission spectra under different dissolved oxygen were recorded at room temperature. Dissolved oxygen was controlled by introducing different ratios of N2 and O2 into the solution within a cuvette, using a pair of mass-flow controllers. We typically allowed 180 s between the changes of the N2 and O2 gas mixtures to ensure a new equilibrium. The solution was excited at a wavelength of 405 nm, and maximum emission was monitored at 660 nm. After using 405-nm light to continuously excite Fe3O4@Os1-PS and Fe3O4@Os1-PMMA for 30 minutes, the intensities of their maximum emissions at 660 remained almost unchanged (Fig. S4), suggesting that these two new kinds of oxygen sensor are particularly stable. It was observed that emission intensities of Fe3O4@Os1-PS (Fig. 3a) and Fe3O4@Os1-PMMA (Fig. 3b) at 660 nm (ascribed to Os1) decreased with increasing O2 partial pressure from 0 % atm to 100 % atm. The emission ratios (I0/I) and oxygen concentrations showed a linear relationship, following the Stern-Volmer Eq. (1).
I0/I=1+Ksv [PO2] Eq. (1)
where I0 and I are the steady-state luminescence signals at 660 nm, measured under nitrogen and under different O2 partial pressures, respectively. KSV is the Stern-Volmer quenching constant, and PO2 denotes the O2 partial pressures. As shown in Fig. 3c, the linear response to oxygen indicates a uniform distribution of the oxygen probes in the PS and PMMA matrix. The sensitivity of the oxygen sensor with nanoparticles is found to be 4.86-fold for Fe3O4@Os1-PS and 2.68-fold for Fe3O4@Os1-PMMA. Fe3O4@Os1-PS presented higher sensitivity for sensing. To explore whether the sensors of Fe3O4@Os1-PS can be reusable for detectiong dissolved oxygen (DO) when they were conserved for some time and redispersed in water, we conducted the extra experiments for the detection of oxygen sensitivity. Interestively, even after a month, Fe3O4@Os1-PS almost retained the same oxygen-detection sensitivity (Fig. S5), indicating that the sensor remained highly stable when redispersed in solution after a long time. Based on the advantages of fluorescence stability and reusable for the detection of dissolved oxygen (DO). Therefore, Fe3O4@Os1-PS was attempted to chosen as the exbacteriaent sensor for oxygen in the following experiment.
The magnetism characterization for Fe3O4@Os1-PS was performed using a vibrating sample magnetometer (VSM) at room temperature. VSM measurements showed that Fe3O4@Os1-PS could be used as a magnetic material because it presents a saturation magnetization (Sm) of 4.59 emu/g (Fig. 4a) which was decreased compared to the pure Fe3O4 nanoparticles due to the fact that small amounts of Fe3O4 was wrapped into the nanoparticles of Fe3O4@Os1-PS. Although the saturation magnetization (Sm) value of Fe3O4@Os1-PS has decreased increasingly, this nanoparticles can be quickly separated by less than 5 s by an external magnetic field. The magnetic nanoparticles may also be recycled using magnets for further use (Fig. 4b).
For biostudies, it is important to determine the toxicity assay for a new material. In this research, E. coli was cultured in LB broth at 37°C, with shaking at 200 rpm for 12 h. We incubated the same E. coli density (OD 0.025) in the multi-well plate with different concentrations of Fe3O4@Os1-PS nanoparticles. Blank 1 was the control test without the sensing nanoparticles, and blank 2 was the control test without E.coli. Absorbance was monitored at a wavelength of 600 nm (OD600) using a plate reader. As shown in Fig. 5a, the time-dependent OD600 of E. coli was not influenced by Fe3O4@Os1-PS concentration, suggesting the non-toxicity and biocompatibility of the nanosensors.
Bacteria respiration is a key process for bacteria metabolism. Bacteria at a higher concentration can consume dissolved oxygen more rapidly. Bacteria were gradually diluted with LB broth to OD600 of 0.1, 0.05, 0.025 and 0.0125. Then 100 µL of the above-mentioned bacteria densities were mixed with 0.025mg/ml of Fe3O4@Os1-PS and placed in 96 well plates to test oxygen respiration, and the same volume of LB broth without bacteria was set as the blank control. To prevent the exchange of oxygen in the media with air, 100 µL of mineral oil was used to seal the well. For an excitation of 405 nm, emission intensity changes at a wavelength of 660 nm were recorded for the oxygen probe during bacteria growth. It was observed that the intensity of the oxygen probes’ emission increased much faster at higher bacteria densities, indicating that the oxygen was consumed much faster at these higher densities (Fig. 5b). After the dissolved oxygen in the medium was completely consumed, the fluorescence intensity did not change further. A slight decrease in the fluorescence intensities at an early stage of the experiments was observed because of the temperature effect. Together, these data showed that the nanosensors could detect the respiration of bacteria, and thus the metabolism of bacteria.